Fads3 modulates docosahexaenoic acid in liver and brain

Highlights

• The first global Fads3 knockout (KO) mice show no overt differences compared to wild type.

• DHA levels on postnatal day 1 (P1) are lower in KO brains.

• DPA to DHA ratio at P1 was lower in KO livers suggesting lower desaturase activity.

• Fads3 may facilitate desaturation while inhibiting elongation during early development.

Abstract

Fatty acid desaturase 3 (FADS3) is the third member of the FADS gene cluster. FADS1 and FADS2 code for enzymes required for highly unsaturated fatty acid (HUFA) biosynthesis, but FADS3 function remains elusive. We generated the first Fads3 knockout (KO) mouse with an aim to characterize its metabolic phenotype and clues to in vivo function. All mice (wild type (WT) and KO) were fed facility rodent chow devoid of HUFA. No differences in overt phenotypes (survival, fertility, growth rate) were observed. Docosahexaenoic acid (DHA, 22:6n-3) levels in the brain of postnatal day 1 (P1) KO mice were lower than the WT (P < 0.05). The ratio of docosapentaenoic acid (DPA, 22:5n-3) to DHA in P1 KO liver was higher than in WT suggesting lower desaturase activity. Concomitantly, 20:4n-6 was lower but its elongation product 22:4n-6 was greater in the liver of P1 KO mice. P1 KO liver Fads1 and Fads2 mRNA levels were significantly downregulated whereas expression levels of elongation of very long chain 2 (Elovl2) and Elovl5 genes were upregulated compared to age-matched WT. No Δ13-desaturation of vaccenic acid was observed in liver or heart in WT mice expressing FADS3 as was reported in vitro. Taken together, the fatty acid compositional results suggest that Fads3 enhances liver-mediated 22:6n-3 synthesis to support brain 22:6n-3 accretion before and during the brain growth spurt.

Introduction

Highly unsaturated fatty acids (HUFA), especially docosahexaenoic acid (DHA, 22:6n-3) and arachidonic acid (AA, 20:4n-6) are metabolically required for growth, early neural and visual development [1], [2], [3]. The perinatal brain growth spurt imposes the highest relative demand for DHA and AA for structural lipid. Rodents are altricial mammals born with immature brains that undergo the brain growth spurt in early postnatal life [4]. The human brain growth spurt is perinatal, starting at about 27 weeks and continuing past 2 years of age. Brain growth and brain size are correlated with neurogenesis [5] during which time HUFA are rapidly accumulated [6].

The availability of HUFA in mammals depends on dietary intake and endogenous synthesis. In an alternating process of desaturation and elongation, HUFA can be biosynthesized from dietary PUFA precursors, linoleic acid (LA, 18:2n-6) and α-linolenic acid (ALA, 18:3n-3). Fatty acid desaturase 1 (Fads1) and fatty acid desaturase 2 (Fads2) code for key multifunctional enzymes in HUFA biosynthesis, introducing cis double bonds at specific position in the carbon chain [7], [8].

The fatty acid desaturase 3 (Fads3) [9] is the third known member of Fads gene cluster, along with Fads1 and Fads2 arose evolutionarily by a gene duplication event and localizes to mouse chromosome 19 [8]. Human FADS homologs with similar structural organization are localized to human chromosome 11q13, a cancer hotspot locus [9], [10]. In human, FADS3 spans 17.9 kb genomic DNA, located 6.0 kb 3′ from FADS2, and has the same gene structure as FADS1 and FADS2 consisting of 12 exons and 11 introns. The amino acid sequences of FADS3 are 52% and 62% homologous to FADS1 and FADS2, respectively. The putative protein coded by FADS3 is composed of an N-terminal cytochrome b5-like domain and three histidine motifs at the C-terminal ends, characteristic of all membrane-bound front end desaturases. FADS3 is transcribed and extensively spliced [11] and the splice variants yield alternative transcripts (ATs) which are phylogenetically conserved in at least several mammalian and avian species [12]. Proteins have been detected that may correspond to ATs [13]. We have shown recently using mouse embryonic fibroblast (MEF) cells and ribosome foot-printing technology the first positive-sequence-specific-proof of Fads3 translation [14]. Despite these observations, there are no reports of Fads3 mediated front-end desaturation.

Existing data suggest that FADS3 has functional significance. Gene expression studies show that Fads3 mRNA changes when Fads1 and Fads2 changes, though not in the same direction. For instance, Fads3 expression increased 3-fold in Fads2 knock-out (KO) mice compared to wild type (WT) [15], and when DHA and AA were included in the diet of neonate baboons, Fads1 and Fads2 expression went down while Fads3 ATs increased [16]. An earlier study showed that the expression of Fads3 was higher in mouse uterus at the implantation site [17]. A recent in vitro study showed FADS3 specificity for trans-vaccenic acid [VA; trans11-18:1], catalyzing putative synthesis of trans11,cis13-CLA isomer (Δ13 desaturation) in the first reported case of a mammalian back-end desaturase [18]. These data, to our knowledge, have not been replicated in vivo. A genome-wide association study (GWAS) showed genetic variants within FADS3 are associated with familial combined hyperlipidemia among Mexican population [19] and a 2011 AVON longitudinal study [20] found minor allele of FADS3 SNP (rs174455) to be negatively associated with DHA in red blood cell (RBC) phospholipids.

Here we present generation of first Fads3 KO mouse colony, with an aim to characterize its metabolic phenotype and to find clues to in vivo function. We focused on the early life because it is the period of rapid brain growth spurt and where intense demand for brain accretion of the most physiologically important HUFA is at least in part mediated by liver.